[LEAPSECS] Leap smear

[mike cook]

Le 24/09/2011 00:00, Clive D.W. Feather a écrit :

> Actually, as we've discussed here ad nauseam, where I live the "day" is

> de jure the mean solar day at Greenwich and de facto 794243384928000 periods

> of the radiation corresponding to the transition between the two hyperfine

> levels of the ground state of the cesium 133 atom.

>

A "day" of MSD length was once in the past the lengh that you state but 
when the SI second  was defined it was already shorter than a mean solar 
second so 86400*SI seconds was already shorter than the MSD. Thus the 
"day" that you want of 86400 SI units has been diverging ever since.

The "day" as indicated by the hands of the clock (which one? ) at NPL 
will be correct if it takes into account the drift.

I do think "day" should be defined as the synodic day, as that is what 
humans without clocks experience. That it varies in length is of no real 
consequence.  As has probably been expressed elsewhere, the controversy 
over the definition of UTC, and the introduction of leap seconds is due 
to what I think is an error in the definition of the second. Since its 
inception people have been trying to measure a rubber band (LOD) with a 
fixed length stick, the SI second. I don't think the SI second should 
have been called a second. It should have been called something like an 
ITU (Intervalle  de temps universal)  and the second left as the rubbery 
86400th of the MSD.
This could be implemented as it was the case in the past. We just need 
in the same transmission the "tick" of TAI so that the requirement of a 
uniform scale is available to those who want one.